Determinants of Employment Growth at MNEs: Evidence from Egypt

DISCUSSION PAPER SERIES
IZA DP No. 1272
Determinants of Employment
Growth at MNEs: Evidence from
Egypt, India, South Africa and Vietnam
Sumon Kumar Bhaumik
Saul Estrin
Klaus Meyer
August 2004
Forschungsinstitut
zur Zukunft der Arbeit
Institute for the Study
of Labor
Determinants of Employment Growth
at MNEs: Evidence from Egypt, India,
South Africa and Vietnam
Sumon Kumar Bhaumik
Queen’s University Belfast
Saul Estrin
London Business School
and IZA Bonn
Klaus Meyer
Copenhagen Business School
Discussion Paper No. 1272
August 2004
IZA
P.O. Box 7240
53072 Bonn
Germany
Phone: +49-228-3894-0
Fax: +49-228-3894-180
Email: [email protected]
Any opinions expressed here are those of the author(s) and not those of the institute. Research
disseminated by IZA may include views on policy, but the institute itself takes no institutional policy
positions.
The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center
and a place of communication between science, politics and business. IZA is an independent nonprofit
company supported by Deutsche Post World Net. The center is associated with the University of Bonn
and offers a stimulating research environment through its research networks, research support, and
visitors and doctoral programs. IZA engages in (i) original and internationally competitive research in
all fields of labor economics, (ii) development of policy concepts, and (iii) dissemination of research
results and concepts to the interested public.
IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion.
Citation of such a paper should account for its provisional character. A revised version may be
available directly from the author.
IZA Discussion Paper No. 1272
August 2004
ABSTRACT
Determinants of Employment Growth at MNEs:
Evidence from Egypt, India, South Africa and Vietnam∗
Foreign investors are expected to contribute to economic development through a variety of
channels. However, many foreign investment operations are small, and almost insignificant in
their impact on the local environment. An important indication of the potential contribution of
foreign investors is thus their employment growth. Employees working for, and trained by, a
multinational enterprise may become carriers of new technology and business practices. The
more employees receive access to new knowledge, the more they in turn may spread the
knowledge across the economy, for instance by setting up their own businesses. In this
paper, we make a first step in investigating the determinants of this important mediating
variable, employment growth. For a dataset covering four diverse emerging economies, we
find that wholly-owned FDI operations have higher employment growth, while local industry
characteristics moderate the growth effect.
JEL Classification:
Keywords:
O13, O33, J21, F23
MNE, employment growth, control, institutions, FDI policy
Corresponding author:
Saul Estrin
Department of Economics
London Business School
Regent’s Park
London NW1 4SH
United Kingdom
Email: [email protected]
∗
The authors are grateful to Simon Commander and Stephen Gelb for helpful comments, to the
Egyptian, Indian, South African and Vietnamese country teams for the collection of data, and to Caitlin
Frost, Maria Bytchkova, Gherardo Girardi for excellent research assistantship. The authors remain
responsible for all remaining errors. The research was made possible by a research grant from the
UK’s Department for International Development.
Determinants of Employment Growth at MNEs:
Evidence from Egypt, India, South Africa and Vietnam
1. Introduction
The literature on foreign direct investment (FDI) has traditionally addressed three
issues, namely, the determinants of the volume of FDI inflow to a country (e.g.,
Borenszstein, de Gregorio and Lee, 1995; Noorbakhsh, Paloni and Youssef, 2001;
Globerman and Shapiro, 2002; Habib and Zurawicki, 2002), the determinants of the
choice of entry mode of a multinational enterprise (MNE) entering a new country
(e.g., Agarwal and Ramaswami, 1992; Hennart and Park, 1993; Görg, 2001; Luo,
2001; Barbosa and Louri, 2002; Gleason, Lee and Mathur, 2002), and the spillover
effects of FDI. The spillover effects, it has been argued, comes largely in the form of
technology transfer by MNCs to their foreign subsidiaries and the consequent
improvement in productivity of domestic firms in the host countries (e.g., Mansfield
and Romeo, 1980; Hasan, 2002; Patibandla and Petersen, 2002, Sinani and Meyer,
2004).
It is now well established that FDI flow to a country increases with improvement in
the quality of its human capital and physical and institutional infrastructure, and with
a decline in factors like policy volatility and corruption. It is also stylised that in the
era of globalized capital markets, where overseas borrowing can be used to
supplement domestic savings, the importance of FDI perhaps lies less in the quantity
of capital inflow than on its ability to transfer technology and business best practices
to the domestic firms in the host country (Findlay, 1978; Borenszstein, de Gregorio
and Lee, 1995). If transfer of technology and business best practices significantly
improves the productivity of domestic firms in the recipient countries, these firms
would improve their international competitiveness, and the impact of this spillover
effect on the economy of the recipient country is arguably much greater than the
impact of the FDI itself. To maximize such benefits to local firms, governments in
many developing countries have stipulated that foreign firms set up business
operations in these countries in the form of joint ventures (JVs), assuming that such
cooperation among multinational enterprises and their local partners would facilitate
the transfer of technology and business practices.
3
However, in the presence of informational asymmetry, MNEs often prefer to use JVs
as vehicles for gaining a better understanding about the business environment in the
host countries, and to develop business relationships with other firms, the
governments and the bureaucracy. Once these relationships are established, JVs are
often dissolved; usually the MNE buys out the equity stake of the others or it enters
the market on its own with a wholly owned subsidiary (see, e.g., Sinha, 2001). It is
easily seen that, given that the expected Nash equilibrium is the dissolution of the JV,
there is no incentive on part of the MNE to transfer technology to its local JV partner,
and this is borne out by empirical evidence (Ramachandran, 1993). Further, MNCs
often use developing countries either as production bases for relatively
unsophisticated inputs for their products, or as target markets for products that are
past their prime in their product cycle (see, e.g., Estrin and Meyer, 2004). Indeed,
there is evidence to suggest that technology transferred to developing countries by
developed country MNEs may be as old as ten years (Mansfield and Romeo, 1980). It
is not surprising, therefore, that the evidence about spillover effects of FDI is mixed.
Some evidence suggests that foreign equity stake in a domestic firm is likely to
improve this firm’s productivity, yet the impact of MNCs’ presence in a developing
country may not necessarily have a productivity-augmenting impact on other
domestic firms, at least not in the same industry (Aitken and Harrison, 1999; Görg
and Strobl, 2001; Meyer 2004).
Despite the barriers to transfer of technology, however, there is evidence to suggest
that some domestic firms in developing countries can become globally competitive by
adopting the state of the art technology, and by adapting to industry best practices.
The Indian software industry is a case in point. In part, this is on account of import of
technology (Vishwasrao and Bosshardt, 2001; Hasan, 2002). However, technology
and business best practices are equally likely to be transferred from MNEs to
domestic firms in developing countries by way of migration of labour from the former
to the latter, a process that is well documented in the context of the Indian software
industry (e.g., Commander, 2003).
The movement of employees is a major element in the transfer of technology within
an industry, and beyond. MNEs build local human capital through training of local
employees, yet these highly skilled individuals may move to local firms or start their
4
own entrepreneurial businesses. Even rank and file employees acquire skills, attitudes
and ideas on the job through exposure to modern organization forms and international
quality standards. Labor mobility can thus enhance productivity throughout the
economy by transferring tacit knowledge that could not be transferred through
informal contacts between firms.
MNEs naturally tend to discourage highly trained employees from leaving by paying
salaries above local standards. Thus labor mobility will be low in emerging economies
where MNEs have substantial advantages over domestic firms. Empirical evidence on
spillovers from labor mobility is far from conclusive. Studies focusing on staff in
MNEs find that most employees that received extensive training stay with their MNE.
For instance, Gershenberg (1987) finds only 16 percent of labor movement from
MNEs to Kenyan firms. On the other hand, studies of successful local firms find that
many entrepreneurs of top managers had prior links to MNEs. For example, Katz
(1987) reports that many managers of local firms in Latin America started their career
with MNE subsidiaries. Altenburg (2000) reports that spin-off electronics companies
in Malaysia maintain close relations as suppliers and subcontractors with the MNE,
while Hill (1982) makes similar observations in the Philippine appliance and
motorcycle industry.
This evidence suggests that the movement of employees may not be large in terms of
the number of individuals moving to local firms. However, those that do leave may
have a substantive impact on the development of indigenous firms, especially if they
set up their own businesses. Such movements may not be against the interest of the
multinational firm if the new firms become business partners, for instance as
suppliers, or by advancing innovations that developed within the sphere of the MNE
but could not be pursued further as they fell outside the core competences of the firm.
The larger the pool of workers that work for MNEs, the greater is the probability of a
significant transfer of technology and business practices from the MNCs to the
domestic firms.1 In other words, even if we make the reasonable assumption that the
1
The potential rate of transfer of technology and business best practices would also depend
on the extent to which the MNEs train their employees. However, data about training offered
to employees at a specific affiliate of a MNE is difficult to obtain. Further, data on training
would not take into account learning by doing in the course of the employees’ regular
5
labourers employed by MNEs comprise a small fraction of the overall labour force in
a developing country (Kraye, Heinrichs and Frobel, 1988), the size and growth rate of
employment at MNCs operating in these countries is important because of the
potential rate of transfer of technology and business practices from these transnational
firms to the domestic firms in these countries, by way of inter-firm labour mobility.2
Despite the importance of direct and indirect contributions of MNE employment for
economic development, we have to date little empirical evidence under what
circumstances MNE create employment in emerging markets. In particular, the impact
of the policy environment and the institutional arrangements on employment creation
remain unexplored. Our contribution to this nascent literature is based on an unique
data set of 293 MNEs operating in four developing countries that differ significantly
with respect to their policy and institutional environment, as also their pool of skilled
labourers, namely, Egypt, India, South Africa and Vietnam. Our results suggest that
availability of appropriate human capital, institutional factors and the extent of control
that a MNE has over its affiliate are important determinants of the growth of
employment of MNEs in emerging markets.
The rest of the paper is structured as follows: In Section 2 we develop the analytical
paradigm underlying the empirical analysis. The data and variables are described in
Section 3. The regression specification and the results are reported and discussed in
Section 4. Section 5 concludes.
2. Labour Demand of MNEs in Developing Economies
Suppose that a firm’s production function is given by
q = f(l, k; m)
[1]
activities. Hence, albeit imperfect, size as measured by the employment level is perhaps the
best available indicator of the potential for technology transfer.
2
Economists have, of course, long argued that a major impact of MNEs on employment in
the host countries is by way of indirect channels like local supply chains (Lall, 1979), and
indeed much of the literature linking FDI with employment growth explores ways to measure
the indirect impact of FDI on employment growth.. But as we have already argued the
importance of MNEs lie in their ability to transfer skills and technology to the local markets
and only the section of the labour force that is directly hired by the MNEs is relevant in this
context.
6
when q is output, l is labour, k is capital, and m is the managerial input (see, e.g.,
Bhaumik and Estrin, 2003). Cost minimisation, which is the dual of profit
maximisation, yields the labour demand function
l = g(w, r, p, q(m))
[2]
when w is the wage rate, r is the rental rate of capital, and p is the price of the final
product. Economic theory suggests that ∂ l/∂ p and ∂ l/∂ q are both positive. In other
words, if there is a growth in sales, whether on account of an increase in p or because
of an increase in q, the demand for labour is likely to increase. In imperfect
competition, the real growth of sales of a firm – the change in q – depends on two
different factors, namely, the growth of the industry to which the firm belongs, and
the market share of the firm within that industry.
Therefore, a MNE’s demand for labour is likely to depend on both the realised and
expected growth rates of the relevant industry in the host country, as well as on its
expected share of the host country market (Watanabe, 1980). The expected rate of
growth, in turn, will depend on the extent to which the industry in the host country is
liberalised, the extent of economic reforms undertaken at the country-wide level. The
expected market share, on the other hand, would depend on the extent of competition
faced by the MNE affiliate in the host country. It would also depend on the ability of
the MNE’s local management to adopt marketing and other strategies that are
consistent with the business environment in the host country, i.e., on the quality of the
managerial input. In any country, but especially in the context of a developing
country, this quality would, in turn, depend on the experience of a MNE in mitigating
the context-specific institutional and other challenges. We discuss some of these
issues below. In sum, therefore, labour demand and employment growth are likely to
be positively related to the industrial growth rate, the extent of economic
reforms/liberalisation both at the country-wide and industry level and the extent of
operating experience of a MNE in the context of developing countries, and is likely to
be negatively related to the extent of competition faced by the MNE in the host
country.
The behaviour of a MNE with respect to its size or level of employment will also
depend on its strategic objectives. MNEs usually set up operations in new countries
either to cater to large local markets, or to gain access to local resources that are
7
valuable in so far as the supply chains of the MNEs are concerned. This distinction
has important implications for how investors set up their operations (Buckley and
Casson 1998). Local production for local markets eliminates the cost of transporting
the product from production locations in other countries, and are also able to eliminate
tariffs from the retail price of the product, thereby making it competitive vis a vis the
domestic and import competition in the new location. Local production facilities also
endow MNEs with the flexibility in the production process that is required to appeal
to local tastes and preferences (Bartlett and Ghoshal, 1989).3 Resource-seeking
investment, on the other hand, permits MNEs to leverage the resources available in
the new production location – be it petroleum in Russia, skilled IT personnel in India
or cheap low-skilled labour in Thailand – to give its global operation a competitive
edge over its rivals. While MNEs tend to be more capital intensive than their domestic
counterparts (Marsh, Newfarmer and Moreira, 1983), it is easily seen that the strategic
objective of a MNE may play a role in determining its optimal labour-capital mix. For
example, it can be argued that a resource-seeking MNE that has entered a developing
country to take advantage of the low wage rate of its skilled (and perhaps semiskilled) labour force is likely to adopt a labour-intensive technology in the host
country,4 whereas a market-seeking MNE might produce its product in the host
country with an unchanged input mix because its strategy is to leverage on its brand to
make super-normal profits in the host country market.
Aside from agency problems, we also have to take into account the fact that in reality
it is usually not possible indefinitely to substitute one factor of production with
another. This is especially true if we view a firm not as a vehicle to transform inputs
like labour and capital into output, but as an organisation that also has to transform the
output into market share, market share into profits, profitability into working capital
etc. In reality, therefore, a mix of several different factors of production like skilled
technical labour, skilled managerial labour, skilled professionals like accountants and
lawyers, and supporting infrastructure in the form of ICT and electricity are required
3
McDonald’s outlets in India, for example, use mutton instead of beef, and their products
have been significantly Indianised to suit the culinary taste of the local consumers.
4
The Hecksher-Ohlin theory about international specialisation argues that most developing
countries have a comparative advantage in labour-intensive products. Therefore, resource
seeking MNEs are even more likely to opt for labour intensive production techniques if they
use a developing country as a location for downstream units of their supply chain, or as an
export base to the rest of the world.
8
for meeting the multiple objectives (or transformation goals) of the firm. In other
words, a firm may not be able to grow in terms of its labour force if it does not have
adequate access to complementary factors like adequate ICT infrastructure.
Further, given the context of emerging markets, we have to take into account paucity
of skilled personnel who can adapt quickly to the MNEs’ technology (McDonald,
Tusselmann and Heise, 2002). Neo-classical theory allows a profit maximising/cost
minimising firm to decide on its optimal labour-capital mix without reference to the
total supply of labour in the factor market. However, this supply side constraint may
be a reality in emerging markets on account of two different reasons. First, the
education system of an emerging market may simply not generate enough labourers
with adequate skills to adapt to the technology and business practices of a MNE. The
concern about skills may not be relevant if the MNE uses its emerging market affiliate
to produce something that is an essential part of its supply chain but one that involves
old technology.5 However, in such a context, any concern about technology transfer is
moot. Second, labourers with adequate skills may be organisationally embedded
and/or geographically fragmented in emerging markets. For example, high skilled
managerial labour may be scarce simply because most skilled managers are owners of
their own firms, and language and cultural barriers in large countries like India may
prevent internal migration of skilled and semi-skilled technical labour. In sum, a
MNE’s growth in terms of employment may be affected by the availability of
labourers with appropriate skills.
The ability of a MNE to grow in an emerging market would also be dependent on the
institutional environment. Clearly, the firm is more likely to grow in size if the
institutional environment – defined by bureaucratic procedures, explicit and implicit
government policies etc – is conducive to conducting business than if the environment
is a hindrance to growth. The quality of local institutions may also impact
employment growth at MNEs indirectly, by way of their impact on the
macroeconomic environment. If weak institutions lead to macroeconomic instability,
it is likely to have a detrimental impact on employment growth at MNEs (Aizenman,
5
Note that such behaviour is consistent with internalisation which is one of the pillars of the
OLI framework (Ethier, 1986).
9
2003). We can, therefore, hypothesise a positive relationship between employment
growth and the quality of the institutional set-up in the host country.
Neoclassical theory that forms the basis for the above analysis assumes that firms are
profit-maximizing or cost-minimising in nature. However, following the seminal
paper by Jensen and Meckling (1976), it is now stylised that a firm’s behaviour is
significantly dependent on the extent to which it resolves the agency problems
involving the various stakeholders of the firm. The agency problem potentially faced
by a MNE involves the divergence between the MNE parent’s objectives and those of
the local management.
This problem is likely to be exacerbated if the MNE parent does not have full control
of the operations of the affiliate, which is the case if the mode of entry into the new
market is a JV or partial acquisition of a local firm (Mansfield and Romeo, 1980;
Ramachandran, 1993). In either of these two cases, foreign and local co-owners have
to find compromises to match their respective objectives for the local firm. Thus, a
MNC would face uncertainty about the extent of convergence between its own
objectives and those of the local management, such that the MNE might be more
reluctant to expand it operations in the host country relative to a situation in which it
has full control of the local operations. Moreover, foreign investors may be concerned
that transfer of technology come to benefit the local partner, who may – in a worst
case scenario – emerge as a competitor (Buckley and Casson, 1998).
We can, therefore, hypothesise that, ceteris paribus, a MNE affiliate is likely to
expand its operations in a host country faster if it has full control over the local
affiliate’s operations than when it has to share control with a domestic firm in the host
country.
3. Data and Variable Measurement
3.1 Survey
The data was collected from randomly selected MNE affiliates operating in Egypt,
India, South Africa and Vietnam, using a survey instrument (see Estrin and Meyer,
2004). The base population for the survey was defined as all registered FDI projects that
were established in the four countries between 1990 and 2000 that had a minimum
10
employment of 10 persons, and minimum of 10 percent equity stake by the foreign
investor. The time limit ensured that the information relevant to the decisions taken at
the time of establishment of these firms was part of the organisational memory at the
time of the survey. Similarly, the stipulations concerning size and equity stake of the
foreign investor ensured that the firms included in the base population were not trading
or sales offices, but rather were fully operational business operations. The questionnaire
was structured to enable us to collect information about not only the characteristics of
the local affiliates, but also about the perception of the affiliates about local conditions
during the recent years of operation. After accounting for missing observations, we
have usable information for 293 observations spread across the four countries.6
3.2 Variable Measurement and Specification
Our measure of employment growth is the average growth rate of the labour force
associated with an MNE affiliate from the inception of its operation in the host
country to 2000. The measures of the explanatory variables are as given below:
We assume that a MNE is resource seeking if it sells less than 50 percent of its output
in the host country’s market, and capture the motivation of the MNE using a dummy
variable that takes the value 1 if a MNE is resource seeking, and zero if it is market
seeking. The availability of local resources is measured for three categories: qualified
personnel, machinery and equipment, and IT and telecommunications services. The
perceptions of the MNE affiliates that responded to our survey are measured on a
Likert scale of 1 (never available) through 5 (readily available). As we have reported
in Tables 2 and 3, we have used dummy variables to reflect these perceptions, with
the dummy variable taking the value 1 if the reported score was 4 or 5 on the Likert
scale.
We have similarly measured eight different aspects of institutional/business
environment in the host country using Likert scale measures that run from 1 (very
conducive to business) to 5 (not conducive at all). Crombach’s alpha suggests that it
would be feasible to group these measures into three different categories, namely,
conduciveness of official procedures, conduciveness of general institutional
6
Of these, 23 percent are from Egypt, 22 from India, 30 percent from South Africa and 25
percent from Vietnam.
11
framework, and conduciveness of government policies. Official procedures include
those associated with obtaining business licenses, and visa and work permits, as well
as those related to real estate purchase and environmental regulations. The general
institutional framework includes the perception of the MNE affiliates about the legal
framework and law enforcement in the host country, and the predictability and
stability of rules and regulations. Finally, the responses to the questions about
government policies take into account the policies of both the central and the local
governments. Using the responses of the MNE affiliates to these eight queries, we
construct indices for conduciveness of official procedures, conduciveness of general
institutional framework, and conduciveness of government policies, when each index
is the average of the responses to the questions included in that category. In part of the
subsequent analysis, we use three dummy variables to capture the quality of business
environment, with each dummy variable taking the value 1 if the value of the
corresponding index is 4 or higher.
We use three different measures of a MNE’s experience in the subsequent analysis,
namely, the number of countries in which the MNE has operations, a dummy variable
that indicates whether or not a MNE had prior experience of operating in a host
country, and a categorical variable that takes the value 1 through 4 depending on the
number of clusters of emerging markets in which the MNE has operating experience.
After some experimentation, we chose four clusters; Africa, Asia (other than Japan),
Central and Eastern Europe, and Latin America.
We measure the characteristics of each industry in each host country using three
variables, namely, the average growth rate of the industry during the 1990s, the
number of competitors of a MNE affiliate in the industry, and the extent of
liberalization and privatization experienced by the industry during the 1990s. In
addition, we use a proxy to capture the dynamics of economy-wide of reforms over
time. Aside from using a categorical variable to measure the extent of local
competition, where 1 stands for none and 5 stands for greater than 10, we also use in
the subsequent analysis a dummy variable that takes the value 1 if the number of local
competitors is greater than 10, and is zero otherwise. The extent of liberalization and
privatization were measured by the four country teams responsible for the survey,
using a Likert scale that runs from 1 (no policy changes) to 5 (major policy changes).
12
We measure the dynamics of economy-wide reforms using a time trend with respect
to 2000, the implicit assumption being that the extent of reforms increased (linearly)
over time. In other words, if the value of the time trend is close to 10 for a MNE
affiliate, it entered the host country early in the 1990s, while if this value is zero then
it entered the host country late in the decade.
We assume that MNE that entered the host countries with Greenfield projects or by
cross-border acquisition of more than 90 percent of a local firm have full control of
the affiliate’s local operations, while those that entered the host country in the form
joint ventures (JVs) and partial acquisition of local firms did not have full control of
the affiliates’ local operations. We, therefore, capture the extent of control a MNC has
over its affiliate’s local operations using a dummy variable that takes the value 1 if the
mode of entry was Greenfield or full acquisition, and zero otherwise.
The econometric specification also includes some control variables. Specifically, we
have controlled for the initial employment level, and the point of entry of a MNE into
a host country. The use of the “time trend” variable to capture the point of entry
allows us to control for the progress of overall economics reforms in the host
countries, under the assumption that the extent of such reforms increased
monotonically over time in all the four countries (see, e.g., Estrin and Meyer, 2004).
The use of trend variables with similar interpretations is stylized in the literature (e.g.,
Bhaumik and Dimova, 2004). We also include dummy variables for the host countries
and the 2-digit industries of the MNEs in these host countries.
3.3 Descriptive Statistics
The descriptive statistics are reported in Table 1. The average growth rate of
employment at the MNE subsidiaries included in the sample was 25.7 percent
between the time of their (post 1990) entry into the host country and the year 2000.
Given that an average MNE entered a host country in 1996 (i.e., six years after 1990),
this translates into around 6 percent growth in employment per annum. This is a small
number given the low initial level of employment of about 181, and the average
output growth of 11.8 percent in the host country industries to which these MNEs
belong. In other words, it is not obvious that high growth of the local industry
13
necessarily translates into employment growth at MNEs. We shall revisit this issue
later in the paper.
Most of the MNEs in the sample, about 71 percent, are market-seeking. However, to
the extent that they require local resources like qualified personnel, ICT and
machinery and equipment, they do not face much difficulty in acquiring them. The
average score for the degree of difficulty in obtaining these resources varies between
3.6 and 4.2, where 1 indicates “never available” and 5 indicates “readily available.”
This suggests that supply side constraints were unlikely to have been binding in the
context of employment growth at these MNEs, perhaps because these firms pay
efficiency wages. The business or institutional environment and the perception about
the extent of liberalization of the host country economies, however, might have posed
a problem, with associated average scores near the middle of the 5-point scale.
Nearly half (45 percent) of the MNEs had commercial experience in the relevant host
country prior to setting up manufacturing operations in it, and about 70 percent of the
MNEs had operational or commercial experience in similar emerging markets.
Indeed, average MNEs in the sample had operational experience in about 25
countries. However, this figure was influenced by MNEs like Coca Cola and Pepsi
that have operations in over 175 countries.
The MNEs faced a modest degree of competition in their host country industries. An
average MNE’s host country industry included 5-10 competitors.
4. Regression Results
The regression results are reported in Table 2. Column 1 reports the coefficients of the
specification in which the measures of resource availability, institutional/business
environment and local competition are formulated as dummy variables. The
continuous/categorical measures of these variables have been used for the
specifications reported in Columns 2-4. Given that the dependent variable for these
specifications is continuous, all specifications reported in Table 2 have been estimated
using ordinary least squares (OLS), with the appropriate correction for
hetereskedasticity. The McFadden’s adjusted R-square estimates for the regressions
are in the range of 0.16 to 0.18, while as well as the F-statistics for the specifications
14
are highly significant and indicate that the specifications are a reasonably good fit for
the data. These statistics are entirely consistent with goodness of fit measures of
cross-sectional regressions involving less than 300 observations.
The results reported in Table 2 indicate that employment growth at a MNE affiliate is
higher if the MNE has full control over the affiliate’s operations. We suggested above
that this might reflect the MNE’s preference for full control before committing
significant resources for growth in a foreign subsidiary in an emerging market. Given
that there were few cases of cross-border takeover in our sample, full control in our
context is near-synonymous with Greenfield projects which add to employment, by
definition. However, we have to note that entry into a host country by way of a
Greenfield project does not necessarily affect the rate of growth of employment.
The negative signs of the coefficients for qualified personnel in IT and
telecommunication services possibly reflect that there is both a quality-quantity tradeoff in employment, and a significant degree of substitutability between labour and
technology-based services. Not surprisingly, employment growth is inversely related
to both the initial number of employees, as implied by Gibrat’s law which suggests
that the size of the labor force in larger firms grow at a slower place relative to the
size of the labor force in smaller firms, as well as to the extent of local competition
faced by the MNE affiliate.
Neither the motivation of the MNE affiliates nor the institutional/business
environment affects the average employment growth during the 1990s. This might be
on account of the possibility that good performance may not be translated into
expansion of operations, especially in a developing country context. MNE affiliates in
developing countries possibly pay efficiency wages, thereby attracting highly
productive employees who may initially operate well within their effort possibility
frontier. Hence, if resource constraints are not binding and if the institutional
environment is favourable, de facto capacity expansion may take place simply by
providing these employees incentives that are compatible with greater effort, without
an increase in the number of employees itself.
15
These lines of argument find support in the results that the rate of employment growth
is negatively associated with the availability of qualified personnel and ICT services,
both of which highlight a quality-quantity trade-off in the context of employment at
these MNEs. If their labour force is qualified, i.e., high-skilled, it is possible for them
to expand their operations, when required, by eliciting greater work effort from the
existing employees and/or by increasing the ICT-labour ratio to increase labour
productivity. A negative impact of an increase in labour productivity, achieved in
either of these two ways, on employment growth is easily envisaged.
Interestingly, even though we have explicitly controlled for experience of the MNEs
regarding operations in the relevant host country as well as in similar emerging
markets, as well as for the extent of economic reforms/liberalization experienced by
the host countries, the timing of entry into the host country clearly matters. Late
entrants add to their labour force much more rapidly than the early entrants. The
possible explanation of this observation lies with the fact that the complex political
economy of economic reforms in emerging markets reaches some steady state only
after some years of intense bargaining among the different stakeholders in the
economy, and hence, rather than leading to a first-mover advantage, early entry into
such a market may lock a MNE into a sector that shows early promise but reform for
which slows down (or even comes to a halt) after some years. For example, although
the banking sector in India was one of the first to be liberalized, the continuing
presence of large state owned banks, and the transformation of large domestic
incumbents like ICICI from development finance institutions to commercial banks led
to a deceleration of the expansion of average foreign banks (Bhaumik and Mukherjee,
2002).
Finally, while employment growth at the MNCs is not sector-dependent, there is a
significant host country effect on such growth. Yet, it appears difficult to explain why
the rate of expansion of a MNE affiliate’s operations is inversely related to the growth
rate of the host-country industry to which it belongs. There are two possible
explanations for this outcome. First, the turnover of local industries is typically
measured in local currencies, which are significantly influenced by exchange rate
fluctuations. For example, the exchange rate for the Indian rupee declined from 35.68
rupees per US dollar in September 1996 (i.e., the time of entry of an average MNE
16
into one of the four host countries) to 46.76 rupees per US dollar by the end of 2000.
However, MNEs are interested in the expansion of their business operations as
measured in the currency of their home base. Hence the greater than 11 percent
growth rate of local industries may not be an accurate reflection of the growth in the
business operations of the MNEs. Second, the growth of turnover in the local
industries may reflect a change in the output mix of these industries, i.e., higher value
addition, which is usually accompanied by a capital-favoring input mix. Note also that
we control via dummy variables for industry affiliation.
5. Concluding Remarks
We have investigated the determinants of employment growth at MNE affiliates in
developing countries. The importance of this line of research lies in the fact that the
level of employment at MNE affiliates is a key determinant of the rate and extent of
transfer of cutting-edge knowledge and business processes from the MNEs to the
domestic firms. The most important finding of our empirical analysis is that
employment growth at MNE affiliates in developing countries is significantly higher
if the MNEs have controlling equity stakes in these affiliates. Aside from highlighting
the importance of control in fostering commitment of the MNE towards its affiliate in
a risky developing country business environment, this result possibly also brings into
focus the technology gap that usually exists between the developed home countries of
the MNEs and the developing countries which they enter. If a MNE has an equity
stake in a developing country affiliate that gives it unambiguous control over
corporate strategies and the associated decisions, it is able to (re)structure the affiliate
such that a transfer of technology is feasible, thereby facilitating the growth of the
affiliate, as manifested by its employment growth.
This result has an important policy implication, namely that FDI policies of
developing countries that aim to foster technology transfer from MNEs to the
domestic firms by limiting the ability of the former to operate wholly owned
subsidiaries in these countries may be counterproductive. MNEs operating in these
countries as minority shareholders in local firms or as JV partners of the latter may
have neither the willingness nor, on account of the technology gap, the ability to
transfer cutting edge knowledge and business processes to the affiliate over which it
does not have complete control. Complete control over the activities of an affiliate
17
enhances the commitment of a MNE towards its affiliate, and enables it to transfer
technology which, in turn, is manifested into growth of employment in the affiliate
and, therefore, into a higher rate of diffusion of cutting edge knowledge and business
processes from the MNE to the local firms. In restricting the equity stake that MNEs
can have in their affiliates, developing country governments ignore the secondary
channels of technology transfer discussed above. This may help to explain why the
empirical literature on FDI finds little evidence of FDI-driven technology
enhancement in developing countries.
18
References
Agarwal, Sanjeev and Sridhar N. Ramaswami (1992) “Choice of Foreign Market
Entry Mode: Impact of Ownership, Location and Internalisation Factors,” Journal of
International Business Studies, 23, 1, pp. 1-27.
Aitken, Brian J. and Anne E. Harrison (1999) “Do Domestic Firms Benefit from
Direct Foreign Investment? Evidence from Venezuela,” American Economic Review,
89, 3, pp. 605-618.
Aizenman, Joshua (2003) “Volatility, Employment and the Patterns of FDI in
Emerging Markets,” Journal of Development Economics, 72, pp. 585-601.
Altenburg, Tilman (2000) “Linkages and Spillovers between Transnational Corporations and
Small and Medium-sized Enterprises in Developing Countries: Opportunities and Policies,”
Proceedings of the UNCTAD X Special Round Table ‘TNC-SME Linkages for
Development,’ Bangkok.
Bartlett, Christopher and Sumantra Ghoshal (1989) “Managing Across Borders: The
Trannational Solution,” Boston: Harvard Business School Press.
Bhaumik, Sumon K and Paramita Mukherjee (2002) “The Indian banking sector: A
commentary,” In: Parthasarathi Banerjee and Frank-Jurgen Richter (eds.) Economic
Institutions in India: Sustainability under Liberalization and Globalization, Palgrave
Macmillan, London, pp. 71-98
Bhaumik, Sumon K. and Saul Estrin (2003) “Transition and Performance: Comparative
Lessons from Chinese and Russian Companies,” Working paper no. 525, Ann Arbor,
Michigan: William Davidson Institute.
Bhaumik, Sumon K and Ralitza D Dimova (2004) “How Important is Ownership in a
Market with Level Playing Field? The Indian Banking Sector Revisited,” Journal of
Comparative Economics, 32, 1, pp. 165-180
Borensztein, Eduardo, Jose de Gregorio and Jong-Wha Lee (1995) “How does
Foreign Direct Investment Affect Economic Growth,” Working paper no. 5057,
National Bureau of Economic Research, Cambridge, Mass.
Buckley, Peter J. and Mark C. Casson (1998) “Analyzing Foreign Market Entry
Strategies: Extending the Internalization Approach,” Journal of International
Business Studies, 29, 3, pp. 539-562.
Commander, Simon (2003) “The Origins and Dynamics of the Software Industry in
Emerging Markets,” Mimeo, CNEM, London Business School.
Estrin, Saul and Klaus E. Meyer (Eds.) (2004) Investment Strategies in Emerging
Markets, Cheltenham: Elgar.
Ethier, Wilfred J. (1986) “The Multinational Firm,” Quarterly Journal of Economics
101, 4, pp. 805-834.
19
Findlay, Ronald (1978) “Relative Backwardness, Direct Foreign Investment, and the
Transfer of Technology: A Simple Dynamic Model,” Quarterly Journal of
Economics, 42, 1, pp. 1-16.
Gershenberg, Irving (1987) “The Training and Spread of Managerial Know-how: A
Comparative Analysis of Multinationals and Other Firms in Kenya,” World
Development, 15, pp. 931-939.
Gleason, Kimberly C., Chun I. Lee and Ike Mathur (2002) “Dimensions of
International Expansions by US Firms to China: Wealth Effects, Mode Selection, and
Firm-Specific Factors,” International Review of Economics and Finance, 11, pp. 139154.
Globerman, Steven and Daniel Shapiro (2002) “Global Foreign Direct Investment
Flows: The Role of Governance Infrastructure,” World Development, 30, 11, pp.
1899-1919.
Görg, Holger (2001) “Analysing Foreign Market Entry: The Choice Between
Greenfield Investment and Acquisitions,” Journal of Economic Studies, 27, 3, pp.
165-181.
Görg, Holger and Eric Strobl (2001) “Multinational Companies and Productivity
Spillovers: A Meta-Analysis with a Test for Publication Bias,” The Economic
Journal, 111, pp. 723-739.
Habib, Mohsin and Leon Zurawicki (2002) “Corruption and Foreign Direct
Investment,” Journal of International Business Studies, 32, 2, pp. 291-307.
Hasan, Rana (2002) “The Impact of Imported and Domestic Technologies on the
Productivity of Firms: Panel Data Evidence from Indian Manufacturing Firms,”
Journal of Development Economics, 69, pp. 23-49.
Hennart, Jean-Francois and Young-Ryeol Park (1993) “Greenfield vs. Acquisition:
The Strategy of Japanese Investors in the United States,” Management Science, 39, 9,
pp. 1054-1070.
Hill, Hal (1982) “Vertical Inter-firm Linkages in LDCs: A Note on the Philippines,”
Oxford Bulletin of Economics and Statistics, 44, pp. 261-71.
Jensen, Michael C. and William H. Meckling (1976) “Theory of the Firm: Managerial
Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics,
3, 4, pp. 305-360.
Katz, Jorge M. (1987) Technology Creation in Latin American Manufacturing
Industries, New York: St. Martin’s Press.
Kreye, Otto, Jürgen Heinrichs and Folker Frobel (1988) “Multinational Enterprises
and Employment,” ILO working paper no. 55.
20
Lall, Sanjaya (1979) “The Indirect Employment Effects of Multinational Enterprises
in Developing Countries,” ILO working paper no. 3.
Luo, Yadong (2001) “Determinants of Entry in An Emerging Economy: A Multilevel
Approach,” Journal of Management Studies, 38, 3, pp. 443-472.
Mansfield, Edwin and Anthony Romeo (1980) “Technology Transfer to Overseas
Subsidiaries by US-Based Firms,” Quarterly Journal of Economics, 95, 4, pp. 737750.
Marsh, Lawrence, Richard Newfarmer and Lino Moreira (1983) “Employment and
Technological Choice of Multinational Enterprises in Developing Countries,” ILO
working paper no. 23.
McDonald, Frank, Heinz J. Tuselmann and Arne Heise (2002) “Foreign Direct
Investment and Employment in the Host Regions,” European Business Review, 14, 1,
pp. 40-50.
Meyer, Klaus E. (2004): Perspectives on Multinational Enterprises in Emerging
Economies; Journal of International Business Studies 35 (4), forthcoming.
Noorbakhsh, Farhad, Alberto Paloni and Ali Youssef (2001) “Human Capital and FDI
Inflows to Developing Countries: New Empirical Analysis,” World Development, 29,
9, pp. 1593-1610.
Patibandla, Murali and Bent Petersen (2002) “Role of Transnational Corporations in
the Evolution of a High-Tech Industry: The Case of India’s Software Industry,” World
Development, 30, 9, pp. 1561-1577.
Ramachandran, Vijaya (1993) “Technology Transfer, Firm Ownership, and
Investment in Human Capital,” Review of Economics and Statistics, 75, 4, pp. 664670.
Sinani, Evis and Klaus E. Meyer (2004): “Identifying Spillovers of Technology
Transfer: The Case of Estonia”, Journal of Comparative Economics 32 (4), forthcoming.
Sinha, Uday B. (2001) “International Joint Venture, Licensing and Buy-Out Under
Asymmetric Information,” Journal of Development Economics, 66, pp. 127-151.
Watanabe, Susumu (1980) “Multinational Enterprises and Employment-Oriented
‘Appropriate’ Technologies in Developing Countries,” ILO working paper no. 14.
21
Table 1
Descriptive Statistics
Mean
Standard
deviation
38.57
Growth rate of employment (percent)
25.69
Ownership and control
Percentage of MNEs with full control over the subsidiary
51.00
50.00
Motivation of MNC
Percentage of MNEs that are resource-seeking
29.00
45.00
Resource availability in host country
3.75
0.90
Availability of qualified personnel⊗
⊗
4.19
0.90
Reliability of IT and telecommunications services
⊗
3.62
1.29
Availability of machinery and equipment
Institutional environment in host country
2.55
0.77
Conduciveness of official procedures⊗ ∗
⊗∗
2.91
0.93
Conduciveness of general institutional framework
2.86
0.97
Conduciveness of policies at various levels of government⊗ ∗
Experience
Number of countries in which MNE present
24.75
37.65
Percentage of MNEs with in country experience
45.00
50.00
Percentage of MNEs with experience in other emerging markets
70.00
46.00
Characteristics of industry in host country
Growth of industry turnover
11.76
11.58
⊗
2.41
0.82
Liberalization and privatization prior to establishment of affiliate
⊗
3.60
1.27
Number of competitors
Control variables
Number of employees at start of operations
181.70
485.61
Year of entry (relative to 1990)
5.85
2.44
Note: 1. ⊗ indicates that the variable was measured using a 5-point Likert scale.
2. ∗ indicates that the Likert scales were inverse, i.e., 1 for “very conducive” and 5
for “not conducive at all”.
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
22
Table 2
Determinants of Performance
(OLS estimates)
Dependent variable: Average growth rate of labour force per annum, from start of
operations until 2000
U
U
Constant
Ownership and control
Full control
Motivation of MNE
Resource-seeking motivation
1
18.52
(15.53)
2
37.87
(21.08)
3
41.71 **
(18.31)
4
40.25
(18.12)
9.75 **
(4.52)
10.58 **
(4.60)
11.51 ***
(4.10)
11.38 ***
(4.15)
2.42
(5.13)
3.82
(5.06)
3.09
(3.98)
Infrastructure sector ×
Resource-seeking motivation
Services (not finance) sector ×
Resource-seeking motivation
Resource availability in host country
Availability
of
qualified - 4.77 *
personnel
(2.46)
Reliability
of
IT
and - 6.30
(5.87)
telecommunications services∇
Availability of machinery and 2.46
(4.63)
equipment∇
Institutional environment in host country
Conduciveness of official 4.82
(4.80)
procedures∇
Conduciveness of general 5.59
(3.31)
institutional framework ∇
Conduciveness of policies at - 6.16
(4.38)
various levels of government∇
Experience
Number of countries in which - 0.23
MNE present
(0.18)
Square of number of countries 0.001
in which MNE present
(0.001)
In country experience
P
P
P
P
P
P
P
P
P
P
Experience in other emerging
markets
Characteristics of industry in host country
Growth of industry turnover
- 0.43 **
(0.18)
Liberalization
and - 2.31
privatization
prior
to (2.90)
establishment of affiliate
- 2.03
Number of competitors∇
(5.27)
Control variables
P
P
3.58
(4.75)
- 8.30
(21.58)
0.052
(9.73)
- 5.35 **
(2.66)
- 4.85 **
(2.42)
2.33
(1.98)
- 6.62 ***
(2.49)
- 3.71*
(2.24)
1.37
(1.84)
- 6.55 ***
(2.48)
- 3.68
(2.26)
1.43
(1.82)
- 3.53
(4.66)
3.04
(3.42)
- 0.30
(2.83)
- 3.59
(4.38)
3.63
(3.16)
- 0.88
(2.45)
- 3.60
(4.37)
3.75
(3.12)
- 0.84
(2.48)
- 1.58
(4.44)
- 0.92
(4.40)
- 1.70
(4.40)
- 0.81
(4.45)
- 0.44 **
(0.19)
- 3.52
(3.00)
- 0.42 **
(0.17)
- 0.64
(2.50)
- 0.43 **
(0.17)
- 0.42
(2.48)
- 1.63
(1.84)
- 3.02 *
(1.62)
- 3.04 *
(1.62)
- 0.24
(0.18)
0.001
(0.001)
23
Number of employees at start - 0.01 ***
of operations
(0.00)
Time trend
4.10 ***
(1.00)
Sector dummies
YES *
Country dummies
YES ***
- 0.01 ***
(0.00)
4.03 ***
(0.93)
YES
- 0.01 ***
(0.00)
3.62 ***
(0.84)
YES
- 0.01 ***
(0.00)
3.61 ***
(0.84)
YES
YES ***
YES ***
YES ***
Mc Fadden’s adjusted R2
0.16
0.17
0.18
0.17
F value
3.93
3.79
3.67
3.48
(Prob>|F|)
(0.00)
(0.00)
(0.00)
(0.00)
N
281
281
293
293
Note: 1. ∇ indicates that the variable was used as a dummy variable in specification 1, and
as a Likert scale measure in the other specifications.
2. The values within parentheses are standard errors.
3. *, ** and *** imply significance at 10%, 5% and 1% level of significance,
respectively.
P
P
24